无菌之门管理-压差报警延迟Part2

接上文：无菌之门管理-压差报警延迟Part1

1. 前言

FDA 7356.xxx 各类中指南中，尤其无菌总会关心一个点，你有没有持续压差监控和管理，压差报警的延迟怎么设定。又或者欧盟无菌附录一中提到的“如果设置了警报延迟，则应在 CCS 内对其进行评估和论证。应定期监测和记录其他气压差。”

当然现阶段，可能有点过于内卷了……

2. 审计发现项

此类发现项，也很多，比如21年，礼来的indi总部的压差报警延迟不合理，一年只有两报警：从一个483发现项看压差报警延时设置来源

又或者：

FDA483：频发的负压倒灌，EMS报警人员瞒报，关键报警无人响应

FDA483:开门后，C级和D级缓冲间频繁负压，长达九个月，无人关心

3. ISPE 月刊

正好，ISPE 的双月刊 制药工程 2024 07月-08月刊里有一篇 《设施设计中的房间压差》，这边分享下里面关于报警延迟的章节，以及我向编委老师（Norman Goldschmidt，ISPE 多本无菌及工艺指南编委，iso14644作者之一，多个美国洁净和无菌标准制定者；Nick Haycocks，同上……）提问的一些回复，基于最近时间不够用，我已经拆成几个part 慢慢写。

本次，高亮部分，是姗姗来迟的来自作者们的一些回复……虽然答案大家心里已有……但是药企么……循证很重要……自己认知的不一定是别人认知的……

增补问题1：跨级别的门打开持续时间内是否需要确认高级别一侧的房间粒子水平始终不超过动态标准？

回答： No – there is no source of particles that would cause a higher particle count.不需要， -没有会导致更高颗粒数的颗粒来源。

增补问题2：假如最坏情况是物品传递过程导致气闸室气流紊乱，粒子超标（动态），在传递结束之后对传递物品和高级别侧区域进行消毒和自净处理，比如自净半个小时后该房间恢复静态粒子标准同时检测物品表面微生物符合高级别侧的标准。这样还研究开门压差报警时间没有什么意义吧？在最坏情况下我们通过清洁消毒和自净处理，确认符合要求后才从高级别一层开门取出物品。

增补问题3：做开门研究以确认压差延迟的时候，是否需要考虑最大活动人数？
回答：yes that would be a worst case scenario 是的，那将是最坏的情况

上一part的内容，在分割线以下

ALARM DELAYS 报警延迟（摘自 ISPE 的双月刊 制药工程 2024 07月-08月 《设施设计中的房间压差》）

Room pressurization, while a regulatory expectation for classified spaces, is not necessarily a primary or process parameter. Rather, it is important as an indication of a space’s ability to protect itself from airborne external contaminants. For this reason, loss of target room pressurization rarely needs to be reported immediately, as loss of pressurization is not necessarily indicative of a loss of clean conditions or incorrect airflow.
房间压差虽然是对分类空间的监管期望，但不一定是主要或工艺参数。相反，它作为空间保护自身免受空气传播外部污染物影响的能力的指标很重要。因此，目标室压差损失很少需要立即报告，因为压差损失并不一定表明清洁条件的损失或气流不正确。

The delay prior to reporting pressurization alarms is commonly set to a few minutes (typically less than 10) and is supported with data on the duration of door opening needed to perform the process and study the impact of the maximum duration of pressure loss. These impact studies may include open-door particle counts, smoke studies of the ingress of air from lower-classification spaces to higher ones, recovery studies from upsets due to door opening, and similar qualification activities to ensure the maintenance of the desired classification when the door is open.
报告压差警报之前的延迟通常设置为几分钟（通常小于 10 分钟），并辅以执行该过程所需的开门持续时间数据，并研究压力损失的最大持续时间的影响。这些影响研究可能包括开门颗粒计数、空气从较低分类空间进入较高分类空间的烟雾研究、因门打开而引起的扰动的恢复研究，以及类似的验证确认活动，以确保在门打开时保持所需的分类。

When establishing the alarm categorization, consider the potential impact of a reverse airflow (room differential pressure reversal or inversion). It is common to focus on internal differential pressure events; however, it is also important to consider external differential pressure events and the potential for ingress from outside spaces. A best practice is to observe the time associated with loss of conditions due to an HVAC system failure to assure that the pressurization alarm delay is set to a safe value before conditions (pressure differentials, particle counts, temperature humidity etc.) are lost. This can be arrived at by simulation or by reviewing historical operational data.
在建立报警分类时，请考虑反向气流（室内压差反转或反转）的潜在影响。通常关注内部压差事件;然而，考虑外部压差事件和从外部空间进入的可能性也很重要。最佳做法是观察由于 HVAC 系统故障而导致的条件损失相关的时间，以确保在条件（压差、颗粒计数、温度湿度等）损失之前将加压警报延迟设置为安全值。这可以通过模拟或查看历史操作数据来实现。

针对月刊注释&回复：

西门问：压差报警延迟是否必须基于开门研究？
a. Normally, yes. The opening of the door typically results in a change in pressurization (depending on room size / HVAC configuration). By doing an “open door study” either with smoke or, preferably, with particle counting gives a scientific basis for how long the door may remain open without significant negative impact to room performance (of the critical room). The amount of time a door may remain open, and the pressure maybe outside the normal range, is set to be less than the amount of time before significant impact.
一般来说，是的。门的打开通常会导致压力变化(取决于房间大小/ HVAC配置)。通过使用烟雾或颗粒计数进行“开门研究”,为门保持打开多长时间而不会对房间性能(关键房间)产生重大负面影响提供了科学依据。门可以保持打开的时间量以及压力可能超出正常范围，被设置为小于显著冲击之前的时间量。

2. 每扇跨区域的门都需要进行报警延迟验证么（C/D,C/B,B/A） ？
a. This is your option, but it is advisable to understand the impact at any room where processing takes place. The exception could be a door between airlocks. 
这是你的选择，但是建议你了解在任何进行处理的房间的影响。气闸之间的门可能是个例外。

3. 开门研究应该使用哪个点会更好，气流流型紊流指出的最差点？日常环境监控点？
a. Great question! The protocol I usually use stipulates 1-2m on each side of the door, assuming that there is no critical activity any closer to the door.However, using your EM points is an excellent choice as well. I would only suggest that you don’t need to do ALL of your room certification points.
很棒的问题！我通常使用的协议规定门两侧各1-2m，假设门附近没有任何关键活动。然而，使用你的EM点也是一个很好的选择。我只建议你不需要做所有的房间认证点。
Note Usually you would complete the room recovery test first following the method described in the ISO Standard – once you have established satisfactory performance you could use a simple local smoke study to determine the direction of airflow when the door is opened – this may help establish the worse case location for the particle counter.
注:通常，您将首先按照ISO标准中描述的方法完成房间恢复测试——一旦您建立了令人满意的性能，您可以使用简单的局部烟雾研究来确定门打开时的气流方向——这可能有助于确定粒子计数器的最坏情况位置。

4. 只测定粒子是否能够作为单一标准（这个问题问的不太好，我会继续问）?
a. Not sure I understand your question. An airlock recovery study (to set a time delay before opening the clean side door) is can be integrated into your “open door” studies, but this is normally separate.
我不太明白你的问题。气闸恢复研究(在打开清洁侧门之前设置一个时间延迟)可以整合到你的“开门”研究中，但这通常是分开的。

5. 参考文章中的“报告增压警报之前的延迟通常设置为几分钟(通常少于10分钟)”，我认为10分钟太长了，肯定会使得房间环境被破坏，一个新的设施没有足够的开门时间数据，那这个时候哪个延迟时间是可取的，2-5分钟？
a. I agree that 10 minutes is too long, in the article I was attempting to quash the idea that the delay must be <1minute. Typical delays in my experience are in the range of 1-3 minutes. It’s hard to justify keeping a door open any longer than that –the time to move a loaded pallet or the EM equipment into an airlock is usually a worst case operation – and that can be timed to make sure the alarm delays wont be triggered by normal operation.
我同意10分钟太长了，在这篇文章中，我试图反驳延迟必须小于1分钟的观点。根据我的经验，典型的延迟在1-3分钟的范围内。很难证明门打开的时间比这更长——将满载的托盘或EM设备移动到气闸室的时间通常是最糟糕的情况——这可以通过定时来确保正常操作不会触发警报延迟。